Bimodality in pan-cancer proteomics reveals new opportunities for biomarker discovery

Bimodal protein expression, characterized by the distribution of protein expression with two modes, is linked to phenotypic variation across various biological systems. Whereas previous studies focused on RNA-seq, we developed a proteomics-specific bimodality model, advancing the identification of cancer biomarkers and targets for precision oncology. We analyzed proteomics data from various cancer types and identified 2401 tumor-associated bimodal proteins which significantly linked to critical cancer pathways, such as amino acid metabolism, extracellular matrix-receptor interaction, and central carbon metabolism. Utilizing an AI-enhanced knowledge graph, we further delineated common patterns among pan-cancer tumor-associated bimodal proteins. A case study on TROP2 in colon adenocarcinoma highlighted up-regulation of MYC and WNT/β-catenin pathways and down-regulation of inflammatory pathways in TROP2-high groups. This highlights the biological differences impacting cancer heterogeneity and vulnerability, aiding treatment decisions. Our findings illustrate the value of proteomics in uncovering novel biomarkers and advancing precision medicine, paving the way for multi-omics integration and clinical validation.